| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 2000, |
| "global_step": 25399, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.00011818933932159319, |
| "grad_norm": 7.239190101623535, |
| "learning_rate": 0.0, |
| "loss": 3.9597, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.23637867864318637, |
| "grad_norm": 4.569056034088135, |
| "learning_rate": 7.088652482269504e-06, |
| "loss": 3.3526, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.23637867864318637, |
| "eval_cosine_accuracy": 0.9351803064346313, |
| "eval_loss": 1.1521742343902588, |
| "eval_runtime": 22.8749, |
| "eval_samples_per_second": 413.423, |
| "eval_steps_per_second": 1.617, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.47275735728637275, |
| "grad_norm": 4.866804599761963, |
| "learning_rate": 1.4180851063829787e-05, |
| "loss": 2.6297, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.47275735728637275, |
| "eval_cosine_accuracy": 0.9370836615562439, |
| "eval_loss": 1.1368498802185059, |
| "eval_runtime": 24.2219, |
| "eval_samples_per_second": 390.433, |
| "eval_steps_per_second": 1.528, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.7091360359295592, |
| "grad_norm": 4.78344202041626, |
| "learning_rate": 2.126950354609929e-05, |
| "loss": 2.27, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.7091360359295592, |
| "eval_cosine_accuracy": 0.9424764513969421, |
| "eval_loss": 1.088667392730713, |
| "eval_runtime": 23.9037, |
| "eval_samples_per_second": 395.63, |
| "eval_steps_per_second": 1.548, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.9455147145727455, |
| "grad_norm": 4.2486572265625, |
| "learning_rate": 2.8358156028368792e-05, |
| "loss": 2.0264, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.9455147145727455, |
| "eval_cosine_accuracy": 0.9423707127571106, |
| "eval_loss": 1.0786563158035278, |
| "eval_runtime": 23.6661, |
| "eval_samples_per_second": 399.601, |
| "eval_steps_per_second": 1.563, |
| "step": 8000 |
| }, |
| { |
| "epoch": 1.181721572794899, |
| "grad_norm": 3.9295907020568848, |
| "learning_rate": 2.9729781514064122e-05, |
| "loss": 1.8456, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.181721572794899, |
| "eval_cosine_accuracy": 0.9449085593223572, |
| "eval_loss": 1.0610214471817017, |
| "eval_runtime": 23.723, |
| "eval_samples_per_second": 398.643, |
| "eval_steps_per_second": 1.56, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.4178769630416814, |
| "grad_norm": 4.169382572174072, |
| "learning_rate": 2.8587197732951956e-05, |
| "loss": 1.7248, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.4178769630416814, |
| "eval_cosine_accuracy": 0.9460716843605042, |
| "eval_loss": 1.0575031042099, |
| "eval_runtime": 23.9067, |
| "eval_samples_per_second": 395.579, |
| "eval_steps_per_second": 1.548, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.6540323532884638, |
| "grad_norm": 4.164876937866211, |
| "learning_rate": 2.6616033219669403e-05, |
| "loss": 1.6326, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.6540323532884638, |
| "eval_cosine_accuracy": 0.9483979940414429, |
| "eval_loss": 1.0424392223358154, |
| "eval_runtime": 23.6735, |
| "eval_samples_per_second": 399.476, |
| "eval_steps_per_second": 1.563, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.8901877435352463, |
| "grad_norm": 3.8136980533599854, |
| "learning_rate": 2.393977593462922e-05, |
| "loss": 1.5739, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.8901877435352463, |
| "eval_cosine_accuracy": 0.9505128264427185, |
| "eval_loss": 1.003242015838623, |
| "eval_runtime": 24.0212, |
| "eval_samples_per_second": 393.694, |
| "eval_steps_per_second": 1.54, |
| "step": 16000 |
| }, |
| { |
| "epoch": 2.1263431337820284, |
| "grad_norm": 3.914799690246582, |
| "learning_rate": 2.071794740754529e-05, |
| "loss": 1.5104, |
| "step": 18000 |
| }, |
| { |
| "epoch": 2.1263431337820284, |
| "eval_cosine_accuracy": 0.9495611786842346, |
| "eval_loss": 1.01349675655365, |
| "eval_runtime": 26.418, |
| "eval_samples_per_second": 357.976, |
| "eval_steps_per_second": 1.401, |
| "step": 18000 |
| }, |
| { |
| "epoch": 2.362498524028811, |
| "grad_norm": 3.88228440284729, |
| "learning_rate": 1.7146081315912403e-05, |
| "loss": 1.4683, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.362498524028811, |
| "eval_cosine_accuracy": 0.9518874883651733, |
| "eval_loss": 0.9978452324867249, |
| "eval_runtime": 24.04, |
| "eval_samples_per_second": 393.387, |
| "eval_steps_per_second": 1.539, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.5986539142755936, |
| "grad_norm": 3.965975761413574, |
| "learning_rate": 1.3445243725964677e-05, |
| "loss": 1.4373, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.5986539142755936, |
| "eval_cosine_accuracy": 0.954002320766449, |
| "eval_loss": 0.9861479997634888, |
| "eval_runtime": 40.0431, |
| "eval_samples_per_second": 236.17, |
| "eval_steps_per_second": 0.924, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.8348093045223757, |
| "grad_norm": 3.817960500717163, |
| "learning_rate": 9.839084604957456e-06, |
| "loss": 1.4122, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.8348093045223757, |
| "eval_cosine_accuracy": 0.9523104429244995, |
| "eval_loss": 0.9923800230026245, |
| "eval_runtime": 25.2947, |
| "eval_samples_per_second": 373.873, |
| "eval_steps_per_second": 1.463, |
| "step": 24000 |
| } |
| ], |
| "logging_steps": 2000, |
| "max_steps": 33844, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 256, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
